Weight-based heparin dosing: clinical response and resource utilization

CLINICAL THERAF’EUTICSVVOL.

20, NO. 4, 1998

Weight-Based Heparin Dosing: Clinical Response and Resource Utilization Cynthia L. Luckie, PharmD, ly2Aileen B. Luzier, PhurmD,1y3 John A. Donovan, RPh 2 Hawari I. Feras, MD,4 and Alan Forrest, PharmDi3 ‘Department of Practice, State Pharmacy, of Pharmacy, 4Department Medicine, Millard

ABSTRACT The objective of this study was to assess a weight-based heparin (WBH) nomogram (SO-U/kg bolus, 18-U/kg-per-hour initial infusion) and determine its clinical performance and impact on resource utilization. All patients treated with heparin for venous thromboembolism or unstable angina during a 15week study period were included in this retrospective, chartreview study. Three groups were identified: patients treated with WBH, patients whose regimen deviated from the weightbased nomogram (DEV), and matched historical controls (HCs). In patients receiving heparin for more than 24 hours, those treated with WBH achieved threshold activated partial thromboplastin time (aPIT) levels significantly faster than did HC or DEV patients. However, 42% of WBH-treated patients were found to have

0149.2918/98/.$19.00

of New at Buffalo Clinical Pharmacokinetics Health Systems, New York

of and

initial supratherapeutic responses. Logistic regression analysis identified age 267 years, prior warfarin therapy within 7 days of heparin, and high initial infusion rate as predictive of a supratherapeutic aPTT response; smoking was predictive of a subtherapeutic response. Bleeding events were not significantly different between groups. An infusion rate of 15 U/kg per hour was found to closely approximate our population’s actual heparin infusion requirement. Resource utilization was significantly different between the WBH and HC groups in terms of nursing interventions at 48 to 72 hours. We concluded that WBH rapidly drives patients’ aPTT response above the therapeutic threshold for heparin; however, prudent adjustment of the initial infusion rate is necessary to avoid a supratberapeutic aPTI response. Our data support a nomogram with an initial infusion rate of 15 U/kg per hour. Key

699

CLINICAL THERAPEUTICS”

words: heparin, administration and dosage; thrombosis; clinical practice nomograms; anticoagulants, administration and dosage; resource utilization.

INTRODUCTION In recent years, much attention has been given to overcoming response variability in patients treated with intravenous heparin. Inadequate anticoagulation during the first 24 hours of therapy in patients treated for thromboembolic or arterial occlusion may increase the risk for recurrent events.‘,* Audits of clinical practice reveal that 50% to 60% of patients fail to achieve a therapeutic activated partial thromboplastin time (aPTT) response within 24 hours. This failure has been attributed to low initial bolus and infusion rates.3,4 Total body weight (TBW) has been identified as a predominant factor in determining a patient’s heparin requirement.2*5 Various weight-based heparin (WBH) nomograms described in recent literature appear superior in achieving therapeutic aPTT responses**“’ relative to standard heparin dosing (typically a 5000-U intravenous bolus followed by an infusion rate of 1000 U/h).9 Anticipated benefits of implementing a WBH nomogram include rapid achievement of a therapeutic aPTT, decreased resource utilization (decreased re-bolus/ infusion changes and laboratory monitoring), and improved quality of care. This study was undertaken to evaluate a newly implemented WBH nomogram for clinical response and resource utilization. In addition, we present predictors of subtherapeutic and supratherapeutic initial aPTI response in clinical practice and a model for identifying a desirable initial infusion rate.

700

PATIENTS

AND METHODS

Study Design This retrospective chart review was designed to evaluate the aPTT response achieved with a WBH nomogram in the treatment of venous thromboembolism and unstable angina. The WBH nomogram used an 80-U/kg bolus with an 18U/kg-per-hour initial infusion rate similar to that proposed by Raschke et al,* with bolus doses rounded to the nearest 1000 U, and initial infusion rates rounded to the nearest 100 U. The nomogram was instituted after a complete educational program was presented to medical residents, nurses, and pharmacy staff. Attending physicians were mailed memorandums detailing the new nomogram. The study was approved by the institution’s internal review board. Patients treated during the implementation of the WBH nomogram whose regimens deviated from the nomogram’s prescribed bolus and infusion rate (DEV) constituted a comparison group. To make a more direct comparison, historical control (HC) patients who were treated with a standard nomogram used at our hospital (a 5000-U bolus followed by a 1000-U/h infusion, with subsequent measurement of aPTT every 6 hours) were obtained from electronic hospital records for the 12 months before initiation of the WBH nomogram.

Patients The 15-week study period began January 1, 1997. Patients treated with intravenous heparin were identified from electronic records coded for WBH or DEV regimens. Male and female patients aged

C.L. LACKIE

ET AL.

218 years with suspected or documented deep vein thrombosis (DVT), pulmonary embolism (PE), or unstable angina were eligible for chart recruitment. All patients were required to have had a baseline aP’IT measurement and complete blood count (CBC) before receiving heparin. Patients receiving concurrent thrombolytic therapy or intravenous nitroglycerin >200 pg/min were excluded, as were patients with antithrombin III deficiency or a history of heparin-induced thrombocytopenia. HCs were matched to WBH patients by age (within 10 years), weight (within 5 kg), sex, and treatment indication for anticoagulation.

Data Collection Primary outcomes of the study were (1) drug performance, which was assessed by time to exceed the therapeutic threshold (aPTT >45 set) and time to attain the therapeutic range (aPTT 45 to 70 set), and (2) resource utilization, which was assessed by counting laboratory, nursing, and physician interventions. Acceptance of the new WBH nomogram was measured by its utilization during the study period. The following data were collected: age, TBW, height, sex, baseline aPTT and CBC, indication for anticoagulation, smoking status, medical history, and current and past medications. Ideal body weight (IBW) was calculated as 50 kg + 2.3 kg X number of inches >60 for men and 45 kg + 2.3 kg X number of inches >60 for women.‘O The times of bolus and/or infusion changes and aPTT blood draws, with respective infusion rates and aPTT values, were recorded over the length of heparin therapy. Dade Actin partial thromboplastin (Baxter Healthcare Corp., Dade Division,

Miami, Florida) of the same lot number was used during both the study period and the HC period. All blood samples for aPTT determination were analyzed using an MLA Electra 1000 automated coagulation machine (Medical Laboratory Automation Inc., Pleasantville, New York). The upper limit of normal used by our laboratory is 30 seconds, and the maximum limit of detection for our assay is 180 seconds. The therapeutic range for heparin at our hospital is 46 to 70 seconds (1.5 to 2.3 X control). In our study, a therapeutic aPTT was defined as two consecutive aPTTs in the range of 46 to 70 seconds. Similarly, an aPTT exceeding the therapeutic threshold was defined as two consecutive aPTTs of >45 seconds. Adverse events such as heparininduced thrombocytopenia and major bleeding were recorded. Heparin-induced thrombocytopenia was defined as a decrease in platelet count, not thought to be the result of other causes, to 2 g/dL, transfusion of two or more units of blood, or hemorrhage in the retroperitoneum, cranium, or a prosthetic joint. Data were collected on nursing interventions, including the number of bolus doses, infusion changes, physician calls, and aPTT blood draws. The number of physician interventions concerning heparin therapy, as evidenced by order entry into the patient chart, was likewise recorded during the periods 0 to 24 hours, 24 to 48 hours, and 48 to 72 hours. The number of times that aPTT was measured during these time periods was also recorded.

701

CLINICAL THERAPEUTICS’

Statistical Analysis Groups were analyzed according to the planned comparisons (WBH versus HC and WBH versus DEV). Continuous variables were analyzed using Kruskal-Wallis analysis of variance. The Bartlett test was used to test for homogeneity of group variances. Dichotomous variables were analyzed using either contingency tables with a chisquare test or the Fisher exact test when the expected value of any cell size was 6. Interval analysis using Kaplan-Meier timeto-event analysis was applied to both the time taken to exceed the therapeutic threshold and the time taken to reach the therapeutic range. The level of significance for all statistical tests was set at P c 0.05. The first aPTI was analyzed using logistic regression (SYSTAT, Systat, Inc., Evanston, Illinois) to identify predictors of initially subtherapeutic (~45 set) and supratherapeutic (>lOO set) responses. Dichotomous independent variables considered were sex, smoking, use of WBH nomogram, obtaining an aPTT before 6 hours, prior treatment with warfarin within 7 days, PE, DVT, unstable angina, comorbidity (eg, malignancy), and use of nonsteroidal anti-inflammatory drugs or antiplatelet medication. Continuous variables, including heparin bolus amount (U/kg), infusion rate (U/kg per hour), and age, were assessed by dividing each variable into quartiles and computing the proportion of aPIT values that were subtherapeutic and supratherapeutic in each quartile. Using recursive partitioning, we determined that age could be coded as a dichotomous variable by patients aged ~67 years and patients aged 267 years. Bolus and infusion appeared to be continuous and progressive in nature, and were entered into the model by respective quartiles. Weight was incor-

702

porated into the model by characterizing patients as being below IBW (120%). Factors that were significantly predictive of the therapeutic infusion rate (the rate at which two consecutive aPTTs were within the therapeutic range) were determined using multiple linear regression with forward stepping. Interpolation was used to determine the infusion that would have given a therapeutic aP’IT in 8 of 11 patients who never achieved therapeutic levels during heparin therapy; in the remaining three patients, a moderate extrapolation (not greater than 15%) was performed to estimate their therapeutic infusion rate. Factors considered for this model were age, baseline aPTT, sex, smoking status, TBW, IBW, percent of IBW, DVT, PE, unstable angina, and malignancy. Not all independent variables were used in any one regression. The regression equation determined from the model, as described above, was used to calculate a predicted therapeutic infusion rate for each patient. The difference between the actual therapeutic infusion rate required by each patient was compared with the calculated infusion rate from the regression equation and with each of six arbitrary initial infusion rates (13 to 18 U/kg per hour). For each of these arbitrary rates, we compared the proportions of patients with infusion rates >50% below (low), 21% to 50% below (moderately low), GO% (adequate), 21% to 50% above (moderately high), and ~50% above (high) their determined therapeutic infusion rate.

RESULTS A total of 81 patients were treated with heparin for DVT, PE, or unstable angina

C.L. LACKIE ET AL.

Table I. Demographic

characteristics

of patients (mean -+ SD, unless otherwise indicated). Patient Group Weight-Based Heparin

Historical Controls

Deviations*

(n = 31)

(n = 31)

(n = 50)

65.5 f 17.1

65.8 f 14.3

66.4 f 14.0

75.3 f 16.7 56-119

75.4 f 16.1 56-115

85.4 f 21.1+ 54-154

Variable Age (Y) Weight (kg) Mean Range Sex (%) Male Female Smokers (%) DVT/PE (%)

15 16 6 16

Unstable angina (%) Baseline aPTT Prior anticoagulation within

15 (48) 25.3 -I-3.2

7 days (%) Other medications NSAIDs Antiplatelet IBW (kg) Mean difference Height (in)

(48) (52) (19) (52)

6 (19)

TBW-IBW

(kg)

4 (13) 10 (32) 62.4 f 9.9 13 f 13 66.8 f 3.7

15 16 9 16

(48) (52) (29) (52)

15 (48) 25.6 + 3.7 2 (6.5) 2 (6.5) 12 (39) 64.8 * 10.8 11 f 13 66.1 f 4.4

24 (48) 24 (48) 11 (22) 28 (56) 22 (44) 25.4 f 3.5 7 (14) 4 (8) 22 (44) 63.1 -c 10.6 22 f 20 66.7 * 3.9

DVT = deep vein thrombosis; PE = pulmonary embolus; aPTT = activated partial thromboplastin NSAIDs = nonsteroidal anti-inflammatory drugs; IBW = ideal body weight; TBW = total body weight. *Patients whose regimens deviated from the bolus and infusion rate prescribed by the nomogram. +P c 0.05 versus weight-based heparin.

during the study (31 WBH and 50 DEV). WBH nomogram use was 46%, 33%, and 30% over weeks 5, 10, and 15, respectively. The baseline demographics of WBH, DEV, and HC groups are shown in Table I. HCs were not found to be different from WBH patients in any of the demographic characteristics studied. DEV patients were significantly heavier than WBH patients (P < 0.05). The range of weights for WBH and DEV patients was 56 to 119 kg and 54 to 154 kg, respectively.

time;

Performance The initial aPTT response for all patients is presented in Table II. WBH patients had significantly higher initial bolus amounts and infusion rates than HC (P < 0.001) and DEV (P < 0.001) patients. No difference was observed in the times of the first aPTT measurement. In the DEV group, three patients had their first aPTI drawn before 4 hours. In these cases, the next aPTI’ time and value, with cor-

703

CLINICAL THERAPEUTICS@

Table II. Initial activated partial thromboplastin

time (aPIT)

response (mean f SD).

Patient Group Weight-Based Heparin (n = 31) Initial bolus (U/kg) Initial infusion (U/kg/h) Time of first aPIT (h) Range No. (%) of patients with aPTT response <45 set 245 set 46-70 set 2100 set

Historical Controls (n = 31)

81.4 f 4.5 17.8 f 0.67 6.8 + 1.7 5-13

69.0 + 13.3+ 15.4 + 2.8+ 6.3 f 2.4 4-16

61.5 f 17.61 13.0 + 3.6+ 6.4 + 1.5 4-11

5 (16) 26 (84) 5 (16) 13 (42)

8 (26) 23 (74) 12 (39)Z 4 (13)”

23 (46)* 27 (54)s 12 (24) 10 (20)s

*Patients whose regimens deviated from the bolus and infusion rate prescribed +P < 0.001 versus weight-based heparin. *P < 0.05 versus weight-based heparin.

responding infusion rate, were considered the first, since aPTTs drawn before 4 hours after the initial bolus and infusion are not reflective of steady state.” The ranges for time of the first aPTT measurement by WBH, HC, and DEV groups were 5 to 13, 4 to 16, and 4 to 11 hours, respectively. The number of patients with initial aPTT response 2100 seconds was 13 (42%), 4 (13%), and 10 (20%) in the WBH, HC, and DEV populations, respectively. The WBH group had a higher proportion of patients with an aPTT 2100 seconds compared with either HC or DEV (P < 0.05) patients. Predictors of a supratherapeutic response were determined to be age 267 years (odds ratio [OR] = 3.78; 95% confidence interval [CI], 1.1 to 12.9; P = 0.03), prior treatment with warfarin within 7 days of hepat-in initiation (OR = 5.93; 95% CI, 1.5 to 23.7; P = O.Ol), an infusion rate of 15

704

Deviations* (n = 50)

by the nomogram.

U/kg per hour to 18 U/kg per hour (OR = 3.94; 95% CI, 1.02 to 15.1; P = 0.05), and an infusion rate of 19 U/kg per hour to 23 U/kg per hour (OR = 11.6; 95% CI, 3.0 to 45.3; P = 0.004). Smoking was the only variable identified as predictive of a subtherapeutic response (OR = 4.1; 95% CI, 1.3 to 13.0; P = 0.02). Twenty-three of 3 1 patients treated with WBH and 41 of 50 DEV patients received therapy for longer than 24 hours. Data for these patients and the respective 23 WBHmatched controls are presented in Table III. Eleven of these 87 patients (13%) (2 [9%] WBH, 5 [22%] HC, and 4 [lo%] DEV) failed to achieve a therapeutic aPTT during heparin therapy. The WBH nomogram produced an aPTT above the therapeutic threshold more rapidly compared with the HC (P < 0.001) and DEV (P c 0.001) patients. At 24 hours, a smaller number of WBH patients had aPITs <45

C.L. LACKIE ET AL.

Table III. Results for patients receiving

heparin for >24 hours. Patient Group Historical Controls (n = 23)

Deviations* (n=41)

81.0 54.7 17.8 f 0.6 7 5-27 22 6-72 15.2 f 3.7

69.4 f 12.9+ 15.4 f 3.0* 33+ 5-101 35+ 6-118 16.0 + 5.0

59.4 f 17.1+ 13.1 f 3.6* 18’ 5-96 24 685 14.1 + 3.0

3 (13) 20 (87) 16 (70)

12 (52)+ 11 (48)+ 9 (39)+

18 (44)+ 23 (56)+ 23 (56)

0 (0) 5.1 + 2.8+

0 (0) 4.3 k 2.0+

Weight-Based Heparin (n = 23) Bolus (U/kg) Infusion (U/kg/h) Median time to threshold aP’IT (h) Range Median time to therapeutic aF’TT (h) Range Therapeutic infusion (U/kg/h) No. (%) of patients with aFlT at 24 h <45 set 245 set 46-70 set 2100 set Length of heparin therapy (d) No. (%) who failed to achieve therapeutic aPTT No. (%) who received warfarin No. (%) with adverse effects Major bleed Thrombocytopenia

2 (9) 3.2 + 1.4 2 (9) 11 (48)

5 (22) 12 (52)

4 (10) 24 (59)

2 (9) 2 (9)

0 (0) 0 (0)

1 (2)

Values are mean f SD. aPlT = activated partial thromboplastin time. *Patients whose regimens deviated from the bolus and infusion rate prescribed +P < 0.05 versus weight-based nomogram. *P < 0.001 versus weight-based nomogram.

seconds compared with either HC or DEV (P < 0.05) patients. A therapeutic aPTT was achieved more rapidly in WBH patients compared with HC (P < 0.05) patients. As a result, more patients in the WBH group were in the therapeutic range at 24 hours compared with HC patients (70% vs 39%; P < 0.05). There was no difference in the proportion of WBH patients with an aPTT response of 2100 seconds compared with either HC or DEV patients at 24 hours.

1 (2)

by the nomogram.

Adverse events were not significantly different between groups (Table III). The two patients who bled in the WBH group were postsurgical patients. One patient was treated with heparin for a DVT diagnosed 3 days postoperatively after a radical mastectomy. Twelve hours after heparin was begun, the patient was noted to have a large hematoma (aPIT = 171 set) around the operative site. The other patient had discontinued heparin for coronary artery bypass grafting. The patient

705

CLINICAL THERAPEUTICS”

Table IV. Laboratory,

physician,

and nursing interventions

(mean + SD). Patient Group

Weight-Based Heparin

Historical Controls

Deviations*

No. of aI?TT measurements O-24 h+

3.4 f 1.0

3.5 f 1.3

3.7 + 0.5+

24-48 h+

3.3 f 0.8

48-72 hs

2.6 f 1.1

3.5 f 0.9 3.3 f 1.0

3.6 + 0.98$

3.5 + 0.6

No. of physician interventions O-24 h+

0.5 f 0.6

0.5 zt 0.8

0.5 f 0.9

2448

h+

0.5 f 0.7

0.4 + 0.7

0.3 + 0.6

48-72 hs

0.1 f 0.3

0.3 f 0.6

0.3 f 0.5

No. of nursing interventions” 624

h+

24-48 h+ 48-72 hs No. of re-boluses

(over 48 h of therapy)

7.6 f 2.2

7.7 f 2.6

8.1 f 1.7

5.1 + 2.1

6.1 rt 2.0

5.3 f 1.5

3.9 f 3.3

5.1 f 1.7$

5.2 f 1.9’

0.7 rf: 0.9

0.6 + 1.1

0.4 k 0.7

aP’Pr = activated partial thromboplastin time. *Patients whose regimens deviated from the bolus and infusion rate prescribed by the nomogram. ‘For O-24 h and 24-48 h, patient numbers were 23,23,

and 41 for weight-based

trols, and patients whose regimens deviated from the nomogram, *P < 0.05 versus weight-based

nomogram.

SFor 48-72 h, patient numbers were 9, 18, and 29 for weight-based tients whose regimens “Bolus, infusion changes,

deviated from the nomogram, physician

heparin patients, historical controls, and pa-

respectively.

calls, aPlT draws.

had returned to a baseline aPTT of 28.5 seconds before surgery but required 4 units of red blood cells immediately postoperatively for a drop in hemoglobin of 7 g/dL. In the DEV group, one patient who was treated with heparin for a PE for 3 days was diagnosed with new-onset melena. Heparin was discontinued, and 2 units of red blood cells were administered for a drop in hemoglobin of 3 g/dL. There were no deaths related to either bleeding events or tbrombocytopenia.

706

heparin patients, historical con-

respectively.

Resource Utilization WBH-treated patients were found to have fewer aP’IT measurements at 0 to 24 and 48 to 72 hours (P < 0.05) (Table IV). The number of physician interventions at 0 to 24 and 48 to 72 hours were not significantly different between the WBH and HC or DEV groups. At 72 hours, WBH patients were found to have fewer nursing interventions than HC or DEV patients (P < 0.05). Although little difference in

C.L. LACKIE ET AL.

resource utilization was observed between the with initial aP’ITs ~100 seconds were found to have an increased number of physician (P = 0.009) and nursing interventions (P = 0.02) in the first 24 hours. Therapeutic Infusion Rate The mean, median, and SD for the therapeutic infusion rate of all 87 patients was 14.9 U/kg per hour, 14.3 U/kg per hour, and 3.8 U/kg per hour, respectively. The therapeutic infusion rate was not different between groups (Table III). Multiple linear regression produced a model with an adjusted R2 of 0.415 as follows: infusion rate = 1234 - (5.4 X age) (21.1 X baseline aPTT) + (10.4 X TBW) The proportions of patients who would receive low, adequate, and high infusion rates given the infusion rate calculated as above, or a set infusion rate between 13 and 18 U/kg per hour, are presented in Table V.

DISCUSSION AND CONCLUSIONS In the present study, WBH patients showed a better aPTT response than HCs in terms of rapidity in exceeding the therapeutic threshold and achieving the tberapeutic range. For patients receiving heparin for more than 24 hours, the WBH nomogram exceeded the therapeutic threshold in a median of 7 hours (compared with 33 hours in the HC patients) and achieved the therapeutic range in a median of 22 hours (compared with 35 hours in the HC patients). Our results are similar to those reported by Shalansky et al* for WBH and control patients. However, in our study unacceptably high numbers of patients had initial supratherapeutic aPITs, which may explain the decline in WBH use at our hospital despite continued in-service presentations during the implementation period. Supratherapeutic aPTT response has been reported in previous WBH studies.2*8,12 Shalansky et al* reported that 36% of patients had initial aPITs of ~100

Table V. Proportions of patients predicted to be above, below, or at their determined apeutic infusion rate relative to actual infusion rate. Dose (Uflrg/h)

Adequate (*20%)

Moderately High (>21%-50%)

ther-

Low (<50%)

Moderately Low (<21%-50%)

13

2

30

58

7

3

14

1

16

70

9

4

15

1

10

66

18

5

16

0

8

58

26

8

17 18

0 0

6 2

46 36

35 49

13 13

Calculated*

0

10

73

16

1

*Infusion rate = 1234 - (5.4 X age) - (21 .l X baseline activated partial tbromboplastin

High (>50%)

time) + (10.4 X total

body weight).

707

CLINICAL THERAPEUTICS”

seconds. In our WBH group, 42% of patients had initial aPTTs of >lOO seconds, with a lo-fold increase in the percentage of initial aPTT values exceeding the upper limit of our laboratory assay compared with HCs (32% vs 3.2%). Shalansky et al8 observed supratherapeutic aPTTs across the patient weight range, with a larger number of supratherapeutic aPTTs in patients weighing >80 kg. As a result, a decrease in heparin bolus and infusion rates for patients weighing >80 kg was proposed. In our study, logistic regression failed to identify TBW, percent IBW, or body surface area as predictors of an initial supratherapeutic aPTT response. In fact, the mean weight of patients with a supratherapeutic response was lower than that of patients who did not have a supratherapeutic response (73 kg vs 82 kg; P = 0.02). In our study, predictors of an initial supratherapeutic aPTT response were age ~67 years, high initial infusion rate/kg, and prior treatment with warfarin within 7 days of heparin treatment. Age has been demonstrated to be an important factor in heparin response. *s5 Some authors have proposed that increasing age may alter coagulation factors, heparin binding sites, or renal clearance of heparin.13 Higher initial infusion rates were observed in patients with initial aPTTs of >lOO seconds. This finding is consistent with the approximated 60minute half-life of heparin (based on a lOO-U/kg bolus dose), implying that the initial bolus would have little effect on a 6-hour aPTT measurement.14 Although prior anticoagulation with warfarin has been used as an exclusion criterion in previous WBH studies, we considered the inclusion of these individuals necessary to determine how WBH would perform in clinical practice, because tar-

708

geted patients often fail anticoagulation treatment. Previous studies have demonstrated the effect of warfarin on aPTT measurements and the increased incidence of supratherapeutic aPTT response with combined warfarin and heparin therapy.15*16This finding is thought to be due to the partial effect of warfarin on the intrinsic coagulation pathway. Smoking was found to be the only positive predictor of subtherapeutic aPTT response (<45 set). This is consistent with the shorter half-life of heparin in smokers treated with intravenous heparin.5v17 In our study population, bleeding complications were not different between groups. Despite controversy over this issue, excessively high aPTT values raise the concern of bleeding.16*18~19Although we are not aware of any WBH study demonstrating increased bleeding events, it would seem advisable to adopt a nomogram that rapidly exceeds the therapeutic threshold without needless overanticoagulation. Previous authors have described therapeutic infusion rates ranging from 13 to 17 U/kg per hour. 2*7-9Using the formula derived from the multiple linear regression equation, a 4.2% error (SD 22.1%) was found compared with the actual infusion rate at which the patient was observed to have achieved therapeutic levels. Thus we believe that use of the multiple linear regression equation in clinical practice would still yield highly variable results and be problematic to institute. Because a standard nomogram is more easily incorporated into clinical use, we carefully examined a range of potential infusion rates. From the predicted distribution of patients in Table V, it is our opinion that a heparin dose of 15 U/kg per hour would be the most appropriate infusion

C.L. LACKIE ET AL.

rate for our population as a whole. The majority of patients would be within 20% of their therapeutic infusion rate, with a minimal number of patients below and a reasonable number (~25%) above that rate. In addition, the distribution of patients with an infusion rate of 15 U/kg per hour is similar to that obtained using the multiple linear regression equation. Other institutions may select a more aggressive rate of 16 U/kg per hour to reduce subtherapeutic responses even further. Generalizability is an important factor in the interpretation of anticoagulation nomograms in clinical studies. We used Dade Actin partial thromboplastin, which is a widely used reagent in the United States. Our study was based on aPTT ratios of 1.5 to 2.3 times the control values that have been proposed by other authors.1,2 Raschke et al2 support using this range when the institution employs this reagent and when the mean control aPTT is ~30 seconds. Some authors have recommended using a therapeutic range corresponding to heparin concentrations of 0.2 to 0.4 U/mL, as determined by protamine titration.20 Currently, antifactor Xa activity has been recommended (0.3 to 0.7 U/mL), although this is not the standard practice in many clinical settings.2’ In addition, using antifactor Xa to calibrate an aPTT range may be imprecise.22 Another limitation of our study is the range of weights studied in the WBH group, which were from 56 to 119 kg. Patients outside this range were not evaluated; thus the results cannot be generalized to all patients. In addition, we did not assess clinical outcomes such as recurrent thrombotic events. Although others have suggested that use of WBH can minimize the number of aPTT laboratory measurements, re-

boluses, and infusion changes, we observed little difference in resource utilization in our patient population. Because supratherapeutic aPTI values resulted in an increase in both physician and nursing interventions, efforts to target patients’ ideal heparin requirements may improve resource utilization further, making WBH use cost-effective.

ACKNOWLEDGMENT This work was presented at the Annual Meeting of the American College of Clinical Pharmacists; November 1997; Phoenix, Arizona.

Address correspondence to: Cynthia L. Lackie, PharmD, Department of Pharmacy, Millard Fillmore Health Systems, 3 Gates Circle, Buffalo, NY 14209.

REFERENCES Hull RD, Raskob GE, Hirsh J, et al. Continuous intravenous heparin compared with intermittent subcutaneous heparin in the initial treatment of proximal-vein thrombosis. NEJM 1986;315:1109-1114. Raschke RA, Reilly BM, Guidry JR, et al. The weight-based dosing nomogram compared with a standard care nomogram. Ann Intern Med. 1993;119:874-881. Wheeler AP, Jaquiss RB, Newman JH. Physician practices in the treatment of pulmonary embolism and deep vein thrombosis.ArchIntemMed.

1988;148:1321-1325.

Fennerty AC, Thomas P, Backhouse G, et al. Audit of control of heparin treatment. Br J Med. 1985:290:27-28.

709

CLINICAL THERAPEUTICS”

5. Cipolle RJ, Seifert RD, Neilan BA, et al. Heparin kinetics: Variables related to disposition and dosage. Clin Pharmacol The,: 1981;29:387-393. 6. Gunnarsson MS, Sawyer WT, Montague D, et al. Appropriate use of heparin. Empiric vs. nomogram-based dosing. Arch Intern Med. 1995:155:52&532. 7. Rivey MP, Peterson JP. Pharmacy-managed, weight-based heparin protocol. Am J Hosp Pharm. 1993;50:279-284. 8. Shalansky KF, Fitzgerald JM, Sunderji R, et al. Comparison of a weight-based hepatin nomogram with traditional heparin dosing to achieve therapeutic anticoagulation. Pharmacotherapy. 1996;16:1076-1084. 9. Hyers TM, Hull RD, Weg JG. Antithrombotic therapy for venous thromboembolic disease. Chest. 1989;95(Suppl):37S-51s. 10. Abernathy DR, Greenblatt DJ. Drug disposition in obese humans. An update. Clin Pharmacokinet. 1986;11:199-213. 11. Cipolle RJ, Rodvold KA. Heparin. In: Evans WE, Schentag JJ, Jusko WJ, eds. Applied Pharmacokinetics. 3rd ed, ~0130. Vancouver, Canada: Applied Therapeutics: 1992: l-39. 12. Nemeth JS, Marxen TL, Piltz GW. Weightbased protocol for improving heparin therapy. Am J Health Syst Pharm. 1996;53: 1164-1166. 13. Campbell NR, Hull RD. Brandt R, et al. Aging and heparin-related bleeding. Arch Intern Med. 1996;156:857-860. 14. De Swart CA, Nijmeyer B, Roelofs JM, et al. Kinetics of intravenously administered

710

heparin in normal humans. Blood. 1982; 60:1251-1258. 15. Hauser VM, Rozek SL. Effect of warfarin on the activated partial thromboplastin time. Drug Intel1 Clin Pharm. 1986;20: 964967. 16. Hull RD, Raskob GE, Rosenbloom D, et al. Optimal therapeutic level of heparin therapy in patients with venous thrombosis.Arch Intern Med. 1992:152:1589-1595. 17. Granger CB, Hirsh J, Califf RM, et al. Activated partial thromboplastin time and outcome after thrombolytic therapy for acute myocardial infarction. Circulation. 1996:93:87&878. 18. Landefeld CS, Beyth RJ. Anticoagulantrelated bleeding: Clinical epidemiology, prediction, and prevention. Am J Med. 1993;95:315-328. 19. Walker AM, Jick H. Predictors of bleeding during heparin therapy. JAMA. 1980; 244:1209-1212. 20. Hirsh J, Raschke R, Warkentin TE, et al. Heparin: Mechanism of action, pharmacokinetics, dosing considerations, monitoring, efficacy and safety. Chest. 1995; 108(Suppl):258S-275s. 21. Schlict JR, Sunyecz L, Weber RJ, et al. Reevaluation of a weight-based heparin dosing nomogram: Is institution-specific modification necessary? Ann Pharmacother 1997;31:1454-1459. 22. Baker BA, Adelman MD, Smith PA, et al. Inability of the activated partial thromboplastin time to predict heparin levels. Arch Intern Med. 1997;157:2475-2479.